Estimating Land Surface Broadband Albedo using Geo-Kompsat-2A/AMI data
Abstract
The land surface albedo, defined as the ratio of solar radiance incident from the sun to land surface and solar radiance reflected from land surface, determines the amount of solar energy absorbed by the Earth's surface. Land surface albedo is one of the importrant climatic factors that determine the energy and moisture balance between land surface and the atmosphere and serves as forcing data for several climate models. Change in land surface albedo are also important indicators of land cover and land-use change, and can help to understrand vegetation-albedo-climate feedback behavior through long-term change analysis. Satellite remote sensing is known as one of the most suitable methods for land surface albedo observations. Especially geostationary satellites observe the same area in a short cycle, so it is possible to observe changes in land surface albedo efficiently. Korea's next generation weather satellite Geo-Kompsat-2A (GK-2A) / Advanced Meteorological Imager (AMI) uses 16 channels to observe the Asia-Oceania region every 10 minutes, making it suitable for continuous land surface albedo estimation and monitoring. The purpose of this study is to produce daily land surface albedo using GK-2A / AMI data. In this study, land surface albedo was estimated through three processing steps including atmospheric correction, Bi-directional Reflectance Distribution Function and Narrow-to-Broadband (N2B) conversion. We performed these processes using 6S radiative transfer model, semi-empirical BRDF model, and linear regression analysis, respectively, and produced black sky and white sky land surface albedo. The estimated land surface albedo showed relative low erroe value of RMSE The GK-2A land surface albedo represented well the general land surface albedo according to the land type (e.g. barren, forest). As a result of verifying with MODIS land surface albedo (MCD43C3), RMSE was 0.027 and 0.029 respectively in black / white sky. The difference between GK-2A and MODIS land surface albedo in the overall area was generally less than 0.05, but GK-2A land surface albedo in some Australian desert areas was somewhat underestimated.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFMGC43I1464L
- Keywords:
-
- 3322 Land/atmosphere interactions;
- ATMOSPHERIC PROCESSES;
- 0480 Remote sensing;
- BIOGEOSCIENCES;
- 1640 Remote sensing;
- GLOBAL CHANGE;
- 1952 Modeling;
- INFORMATICS